Phosphate Recovery by High Flux Low Pressure Multilayer Membranes

Article pubs.acs.org/Langmuir

Phosphate Recovery by High Flux Low Pressure Multilayer Membranes V. J. Disha,† C. T. Aravindakumar,‡,§ and Usha K. Aravind*,∥,⊥ †

School of Chemical Sciences, ‡School of Environmental Sciences, §Inter University Instrumentation Centre, ∥Advanced Centre of Environment Studies and Sustainable Development, Mahatma Gandhi University, Kottayam 686560, Kerala, India ⊥ Centre for Environment Education and Technology, Kottayam 686008, Kerala, India S Supporting Information *

ABSTRACT: This work illustrates the potential use of PEI/ PSS bilayers assembled via layer by layer (lbl) method on a nylon microfiltration membrane for the recovery of phosphate from water in the presence of chloride under ultrafiltration conditions. A total of nine bilayers were used for the selective recovery of phosphate. Bilayers were constructed from polyelectrolyte solutions of varying ionic strength (0−1 M of NaCl). The selected pH for deposition of PEI (5.9) and the presence of supporting salt in the polyelectrolyte solution is expected to provide membranes with high permeability and high charge density. This particular combination of bilayers yielded high flux membranes that allowed selective removal of H2PO4− in the presence of Cl− at low pressure (0.28 bar). The magnitude of negative solute rejection of chloride showed increasing trend with the number of bilayer for a particular salt concentration. Whereas the increase in magnitude with ionic strength is so high (−6.18 to −269.17 at 0.5 M NaCl for 9 bl) that gave the best observed Cl−/H2PO4− selectivity (310.23, flux 13.53 m3/m2-day). To the best of our knowledge, this is the first time a multilayer polyelectrolyte system with such a high selectivity and rejection for H2PO4− is reported. The solution flux decreased with the number of bilayers and ionic strength. The rejection of phosphate was dependent on feed pH, concentration of deposited polyelectrolyte solution, and composition of membrane support.

1. INTRODUCTION The subject of phosphate (P) recovery from various sources has been in the forefront in recent years since its natural resources are being depleted at an alarming rate, raising a threat to the agricultural sector.1−5 On the other hand eutrophication is a leading environmental issue for aquatic sources, the reduction of which is desired. Nonrenewability of phosphate mines, geological imbalances, and a steep increase in fertilizer price underline the significance of the development of a protocol for its recovery.6 Overfertilization together with modern agricultural practices leads to excess runoff of nutrients into water bodies eventually causing dead zones in oceans in an everincreasing in number.7 The global P-flow analysis indicates that two sectors represent the major flow into surface water: nonpoint-source soil erosion and runoff that has relatively low P concentrations and high volumetric flow and the flow of animal waste with high P content and low volumetric flow. Once it reaches the water source, it can easily settle together with the particulate matter making recovery almost impossible. Rittman et al. recently summarized the different methods adapted for the recovery of inorganic P from animal waste.8 The P removal/recovery technologies from aqueous streams still have many draw backs, the main drawback being energy intensiveness. © 2012 American Chemical Society

Ultrafiltration (UF) is a low pressure driven membrane separation method used for the removal of inorganic micropollutants.9 Membrane separation methods which include nanofiltration (NF) and UF are the most preferred physical processes for the treatment of wastewater. The appeal of membrane separation is that it can be conveniently incorporated into a process and separation can be carried out at room temperature. These methods look even more versatile today because of the use of nanostructured membranes fabricated via layer by layer (lbl) assembly of polyelectrolytes (PE).10 It affords a porous support and a self-assembled skin. The skin thickness mainly depends on deposition parameters. Separation characteristics of the membrane can be formulated by controlling the charge density, hydrophilicity, and pore size (composition, pH, supporting salt, etc.). Ion transport through polyelectrolyte multilayer membranes using nanofiltration and a diffusion dialysis mode has been extensively discussed by Bruening et al.11−14 Hong et al. recently reported selective separation of phosphate/chloride using PSS/PDADMAC modified alumina membranes with a rejection rate of 98% Received: June 10, 2012 Revised: August 2, 2012 Published: August 7, 2012 12744

dx.doi.org/10.1021/la302347k | Langmuir 2012, 28, 12744−12752

Langmuir

Article

(NF, flux, 2.4 m3/m2-day at 4.8 bar, selectivity 48).15 Selective removal of phosphate in the presence of anions was also discussed by Ahmadiannaminia et al. with a rejection rate of 94.7% (NF, normalized flux, 0.006 m3/m2-day at 6 bar, selectivity 2.68).16 The nanoporous structure provides a platform for the separation of compounds with a broad molecular weight range under pressure driven convective flow.17−20 The lbl assembled nanostructured material possesses a high surface area to volume ratio and ensures rapid access of analyte molecules to active sites. Filtration methods are energy demanding, but this can be turned down by proper selection of substrate and coating materials.15 Here we propose a low energy route for the effective separation of a phosphate/chloride system from aqueous solution at a low concentration range (100 ppm). This study investigates the possibility of the use of a composite membrane under UF condition for the selective removal of phosphate in the presence of chloride, developed by deposition of a few bilayers of polyethyleneimine (PEI) and polystyrene sulfonate (PSS) on porous polymer support. PEI obviously is an ideal choice as a cationic polyelectrolyte. It has a high capacity for anion binding (especially phosphate) through electrostatic and hydrogen bonding.21 PEI and PSS (weak/strong) together at pH ∼6 form highly permeable layers.22 We have also made use of the property of salt addition to the depositing medium. This would yield less interpenetrating, highly mobile chains with more ionic mobility. Salt addition to the polyelectrolyte solution is also expected to increase the porosity as they form a coiled structure.23 Modified membrane support in ion separation has gained momentum in water purification as they possess high rejection capacity. The low pressure driven membrane separation process has an upper hand in water purification with the introduction of thin film assemblies on conventional ultrafiltration/microfiltration membranes.24,25 Earlier we reported the possibility of modified microfiltration membranes in the separation/recovery of a wide spectrum of compounds (∼67 kDa to 110 Da) from aqueous solutions as well as from effluents.17−20 The flux and permeation characteristics could be well modulated by the stimulation of polyelectrolyte pair, concentration of permeated species, pH of the deposited and filtered medium, and ionic strength. This triggered the possibility of such membranes for the selective recovery/removal of anions like phosphate from water.

of the membranes were kept in water until use after polyelectrolyte deposition. 2.3. Film Characterization. ATR-FTIR (Shimadzu IR Prestige21) of bare and multilayer membranes were obtained using 40 scans at 4 cm−1 resolution, and the spectrum of Zn−Se crystal in air served as a background. A scanning electron microscope was used for the characterization of the multilayer deposited on the bare membrane. Samples were fractured in liquid nitrogen and sputter coated with platinum using JEOL JFC-1600 autofine coater. Images of cross sections and lbl coated surfaces were taken. The EDS spectrum of samples after ultrafiltration was taken using a JEOL-JSM 6390 SEM equipped with swift ED-EDS from oxford instruments. 2.4. Transport Studies. Filtration experiments were carried out using an Amicon 8050 dead end ultrafiltration cell (0.28 bar, 600 rpm) at room temperature. The concentrations of NaCl and NaH2PO4 were 100 ppm. For the filtration experiments with nitrate/phosphate and humic acid/phosphate, the same feed concentrations were used. Anion concentrations were determined using a Dionex 1100 ion chromatograph with an Ionopac AS12A column and a conductivity detector. The eluent was 2.7 mM Na2CO3/1.0 mM NaHCO3 at a flow rate of 1.0 mL/min. Humic acid was detected using a UV−vis spectrophotometer (Shimadzu) at 272 nm. Percent rejection, R, is defined in eq 1 where Cperm and Cfeed are the solute concentrations in the permeate and feed, respectively. The selectivity, S, for solute A over solute B is defined in eq 2, which can be expressed in terms of rejections as shown R(%) = 1 −

S=

C perm Cfeed

CA,perm C B,feed CA,feed C B,perm

× 100

=

100 − RA 100 − RB

(1)

(2)

The permeate solution flux, J, was calculated using eq 3 where V is the volume of permeate collected, A is the effective membrane area, and Δt is the sampling time

J=

V AΔt

(3)

For stability analysis of bilayers, membranes were kept in NaCl (3 M and 5M) solutions for 24 h.

3. RESULTS AND DISCUSSION 3.1. ATR-FTIR. The formation of bilayers was analyzed using sulfonate peaks as the indicator infrared band. In either case, with the increase in number of bilayers or increase in supporting salt concentration, the peak height increased correspondingly (Figures 1 and 2). The peaks at 1035 and 1008 cm−1 are attributed to stretching (SO3−1) and ring vibration of PSS, respectively. The support has no absorption at this region. A similar pattern was observed for membranes with other numbers of bilayers. In Figure 3 the intensity of peak at 1035 cm−1 is plotted as a function of the number of bilayers. The plot shows linear growth, where an equal amount of polyelectrolyte is adsorbed at each step with exact charge compensation until a salt concentration of 0.5 M. At higher salt concentration, exponential growth is observed leading to higher thicknesses. The exponential growth of films to micrometer thickness has been discussed by several groups both experimentally and theoretically.26,27 This growth regime is attributed to the in and out diffusion of one or both of the polyelectrolytes involved. However, in this case the diffusion of PEI seems to be more probable.27 In the case of exponential growth, diffusion of the involved polyelectrolyte occurs throughout the film leading to films of micrometer thickness. 3.2. SEM. A cross sectional view of a 10 bilayered membrane is presented in Figure 4. The skin layer formed seems to have good contact with the porous structure with the pores still

2. EXPERIMENTAL SECTION 2.1. Materials. The supporting membranes used were nylon 6,6 (0.45 μm pore size, Pall Life Sciences) and Supor 450 (0.45 μm pore size, polyethersulfone, Pall Life Sciences). Poly(ethyleneimine), (PEI, Mw = 750 000, 50 wt % solution in water, Sigma Aldrich), poly(styrene sulfonate) (PSS, Mw = 70 000, 30 wt % in water, Sigma Aldrich), NaCl (Sigma Aldrich, ACS reagent grade), NaH2PO4 (Merck), and humic acid (Sigma Aldrich) were used as received. Deionized water (Milli-Q, 18.2 MΩcm) was used for membrane rinsing and preparation of polyelectrolyte and feed solutions. 2.2. Film Deposition. The supporting membranes were rinsed with deionized water and kept in water for 24 h. For the sequential adsorption of the layers, the supporting membrane was first immersed in the solution of the cationic polyelectrolyte (PEI, 0.02 M, pH 5.9), rinsed with deionized water, then immersed in the solution of the anionic polyelectrolyte (PSS, 0.02 M), followed by rinsing with water. The procedure was repeated until the required numbers of bilayers were formed. Immersion time in the individual polyelectrolyte solution was kept as 15 min. The pH of polyelectrolyte solutions were adjusted with 1 M HCl. Bilayers with different salt concentrations were deposited by varying the NaCl concentrations in the range 0−1 M. All 12745


Langmuir

Article

3.3. Single Salt Ultrafiltration. This section first describes the results obtained for H2PO4−/Cl− transport through PEI/ PSS at a solution pH of 5.4. At this pH the predominant phosphate species is H2PO4−.15 The transport profile along with the selectivity is described with respect to membranes developed from different salt concentrations. Comparison of transport characteristics is analyzed in terms of substrate, concentration of polyelectrolyte and pH of feed solution. Single salt ultrafiltration experiments were performed using different bilayers of PEI/PSS membranes prepared from polyelectrolyte solution containing 0.5 M NaCl. Feed solutions containing 100 ppm each of NaCl and NaH2PO4 were used for the filtrations (see the Supporting Information). The rejection of phosphate ion increases with the number of bilayers for single salt ultrafiltration through multilayer membranes. From six bilayers onward, an appreciable rejection of phosphate was observed, and it increases to 58.28% for 9 bilayer. As the bilayer number increases, the permeate flux decreases (it is inversely proportional to film thickness) as expected due to the increase in film thickness. The difference in the Stokes radii of chloride and H2PO4− (Table 1) in principle aids size based separation. The most interesting aspect observed is the negative solute rejection of chloride for the filtration of NaCl. Negative solute rejection is observed mostly in UF membranes of large pore size. This occurs due to the differences in streaming potential that in turn disturb the net current flow to zero.9 However, this phenomenon is normally observed in mixed ion filtration. So the introduction of a second salt (NaH2PO4) is expected to increase the permeate enrichment of chloride that would then selectively filter out phosphate. 3.4. Binary Salt Ultrafiltration. The percent rejection values and selectivities of H2PO4−/Cl− transport through the membrane prepared from polyelectrolyte solutions of varying ionic strength (0−1 M of NaCl) under ultrafiltration condition are presented in Table 2. PSS was at its native pH whereas the pH of PEI was maintained at 5.96 to have high charge density. The percent rejection of phosphate and chloride varied with the number of bilayers and with the added salt in PE solution. The results are given from six bilayers onward. At six bl, the membrane showed a rejection of 77.6% for a salt concentration of 1 M in PE solution for H2PO4−. The variation in rejection percent was more appreciable with increments in salt than those in bilayer number. The selectivities also showed a more or less similar trend. Even in the absence of supporting electrolyte, rejection percent of H2PO4− increased from 6.1 for (PEI/PSS)6 to 32.9 for (PEI/PSS)9, whereas for Cl− there is negative solute rejection (permeate enrichment). Permeate enrichment is observed in cases where the concentration of Na+ in the mixed feed solutions is greater than that of Cl−. With 1.0 and 0.5 M NaCl as supporting electrolytes there was not much difference in the rejection percentage of H2PO4−. At 0.5 M NaCl as the supporting electrolyte, the phosphate rejection increases from 69.7% for (PEI/PSS)6 to 98.8% for (PEI/PSS)9, and correspondingly the Cl−/H2PO4− selectivity increases from 8.3 to 310.23. The energy dispersion analysis (EDS) spectrum of phosphate adsorbed 8 bilayer membranes is given in the Supporting Information, which indicates the presence of phosphate in the membrane. From Table 2 it is clear that even the membranes, prepared without the addition of salt, rejected H2PO4−. This is logical since the Stokes radii of H2PO4− is comparatively higher than chloride. At the same time chloride showed permeate

Figure 1. FTIR spectra of PEI/PSS films consisting of different numbers of bilayers (prepared from a 1.0 M NaCl containing dipping solution).

Figure 2. FTIR spectra of PEI/PSS films consisting of seven bilayers prepared from different salt concentrations in the dipping solution.

Figure 3. Intensity of the sulfonate peak at 1035 cm−1 as a function of the number of bilayers at different salt concentrations.

open. An SEM image of eight bilayers prepared from different salt concentrations (see the Supporting Information) revealed the formation of thick polyelectrolyte bilayers on the support. The bilayers looked more flat and smooth. 12746


Langmuir

Article

Figure 4. (A) SEM cross section of 10 bilayered PEI/PSS (0.5 M NaCl) on nylon. SEM image of (B) unmodified nylon and (C) eight bilayered PEIPSS in 0.5 M NaCl.

multilayers is maintained by intrinsic charge compensation. Charge is balanced by polymer segment counterions. At each step of deposition, counterions are displaced by oppositely charged polyelectrolyte units. This phenomenon is discussed extensively in literature.30−33 That is, in the absence of added salt in PE solution, the multilayer is free of salt in bulk. As we introduce salt into the depositing medium, charge neutrality is attained by extrinsic compensation.13,34 Here polyelectrolyte pairs are charge compensated by salt ions in the depositing medium. In either case, intrinsic or extrinsic compensation ion exchange takes place among the polymer ion pair and salt ions. As more and more salt is added, the extrinsic compensation increases and causes more swelling. This would create highly permeable porous layers. This is exactly the reason that leads to the high selectivity observed for the H2PO4−/Cl− pair at the ninth bilayer for a build up salt concentration of 0.5 M (310.23, Table 2). In this experimental regime, perhaps the ideal condition is achieved for the PEI/PSS pair for maximum

Table 1. Molecular Weights (Mw), Stokes Radii (rs), and Aqueous Diffusion Coefficients (D) of Anions42 anions

Mw (g mol−1)

rs (nm)

HPO42− H2PO4− −

96.0 97.0 35.5 62.0

0.323 0.256 0.121 0.129

Cl NO3−

D (m2/s) 0.76 0.96 2.03 1.90

× × × ×

10−9 10−9 10−9 10−9

enrichment which increased remarkably with salt concentration. This immediately point out the construction difference of bilayers in the presence of added salt. The presence of salt in the depositing solution yields thicker, less interpenetrating, and open membranes with excess charge spread on the surface.16,28,29 Because of high surface charge build up on the surface, more polymer material is adsorbed on each deposition step. This leads to an unusually large increased thickness with added salt. In the absence of salt, charge neutrality in

Table 2. Percent Rejections and Selectivities of Binary Solution UF of Cl−/H2PO4− at pH 5.4 as a Function of Bilayer Number and NaCl Concentration in the Dipping Solution (0−1.0 M)a chloride rejection (%)

Cl−/H2PO4− selectivity

phosphate rejection (%)

NaCl conc. (M)

6 bl

7 bl

8 bl

9 bl

6 bl

7 bl

8 bl

9 bl

6 bl

7 bl

8 bl

9 bl

0 0.1 0.2 0.3 0.4 0.5 1.0

−1.08 −7.38 −27.18 −76.81 −99.57 −151.08 −183.21

−3.49 −2.09 −63.08 −106.33 −142.12 −212.34 −220.14

−6.70 −14.52 −70.82 −114.03 −170.33 −219.25 −242.31

−6.19 −27.39 −146.04 −137.24 −209.47 −269.17 −298.25

6.06 4.59 25.82 53.11 49.94 69.74 77.56

13.46 12.63 37.46 51.13 66.78 81.89 84.76

28.35 39.34 53.25 55.37 69.74 96.58 94.63

32.88 39.61 60.64 70.66 71.10 98.81 97.96

1.08 1.13 1.71 3.77 3.99 8.30 12.62

1.20 1.17 2.61 4.22 7.29 17.25 21.01

1.49 1.89 3.65 4.80 8.93 93.35 63.75

1.58 2.11 6.25 8.09 10.71 310.23 195.22

a

% rejection of Cl− through unmodified nylon = −3.89; % rejection of H2PO4− through unmodified nylon = −2.125. 12747


Langmuir

Article

Scheme 1. Schematic Representation of Transport of Cl−/H2PO4− through PEI/PSS Multilayers

Three transport processes contribute to the ion flux: convective, diffusive, and electromobility. In a high Peclet number regime, diffusional ionic flux can be neglected as the membrane has a large pore size. In the above equation, only convective and electromigration are taken into account. In a pressure driven condition, the initial uneven distribution of coand counterions causes streaming potential which in turn results in electromigration. The streaming potential would rise until zero current flow is reached; however, in the case of the polyelectrolyte multilayer the point of zero current flow, which in fact determines the electrodynamic properties of the membrane, is affected by the pH and ionic strength of the electrolyte and the structure of the multilayer.36 The structure of the multilayer is very much dependent on the construction parameters. In the present system, the ionic strength of the electrolyte is varied. The zero-point streaming current does depend on the electrolyte concentration as per the reports of Duval et al.36 This report further illustrates that the zero-point streaming current represents the overall hydrodynamic, electrostatic, and structural properties of backbone layers. This means that for multilayers the streaming potential generated to bring the net current flow to zero depends on the constituting bilayers. The layers are alternately charged and the co-ions are different (Cl− and H2PO4−) leading to a change in flux of convection and electromigration. As the co-ions here have different mobilities (the chloride ion is more mobile and is smaller than H2PO4−) negative solute rejection of the more mobile co-ion takes place.37 The presence of the less permeable ion, H2PO4−, makes this negative rejection of chloride even more. To further support the negative solute rejection of chloride, an earlier explanation of availability of ion exchange sites with increment in build up salt concentration can be taken into account. At low salt concentration, bulk ion exchange sites cannot be expected as the bulk of the multilayer is normally intrinsically charge compensated. The added salt increases the availability of ion exchange sites. The diffusion of ions through such ion exchange sites can be explained by the hopping model proposed by Schlenoff et al.31 In PES/PSS multilayers we have extrinsically compensated layers and excess negative charge on the surface. When such a membrane is exposed to the feed solution (NaH2PO4 + NaCl), ion exchange takes place for H2PO4−. This would lead to the

thickness with large ion exchange sites. Beyond this particular ionic strength, inadequacy in interaction may slowly increase, and beyond a salt concentration of 1 M, salt induced dissociation might take place. Probably we are approaching this transition at the eighth and ninth bilayers for a salt concentration of 1 M where the observed selectivity is low compared to that for 0.5 M NaCl in PE solution. The separation of charged species through membranes can be charge based, size based, or both. Since PSS is the terminating layer, excess negative charge is present on the surface which can cause the electrostatic repulsion of incoming anionic species. Further support for the rejection of H2PO4− is obtained from size exclusion. Both charge and size factor seem to be the reason for the observed removal of phosphate from the feed. Moreover the underneath layer is PEI which has a high affinity for phosphate ions. These factors would together favor the removal of phosphate species from the feed. There is greater concentration of Na+ in the feed than counteranions. As Na+ ions pass through the PEM, more Cl− ions are dragged along with them to maintain electroneutrality. The diffusion potential is created due to the difference in concentration of sodium and chloride ions in the solution which can enhance chloride transport.13 Negative rejection of electrolytes is reported for pressure driven membrane separation processes using reverse osmosis, nanofiltration, and ultrafiltration membranes. The possible mechanism for this phenomenon depends on several factors that include the large pore size and the difference in the mobility of co-ions. The flux of a binary electrolyte mixture across a charged membrane under pressure driven transport can be described by Nernst− Planck equation given below.9,35 ⎛ dϕ ⎞ ji = Cm,i Γi⎜Jv + Z iDi ⎟ ⎝ dx ⎠

(4)

where ji and Jv stand for the ion and volume flux, Cm,i is the ion concentration at the membrane surface, Γi is the ion distribution coefficient, Di is the diffusivity of the ion, and zi is its valency. The streaming potential generated to give zero current flow is represented by dϕ/dx. The point of zero streaming current for the polyelectrolyte multilayer is defined as the pH where the streaming current is zero which identifies the zero streaming potential.36 12748


Langmuir

Article

presence of Cl− and H2PO4− anions which would then compete for the positively charged polymer unit as shown in eq 5. Pol+Cl−(m) + H 2PO4 − ⇌ Pol+H 2PO4 − + Cl−

Table 5. Solution Fluxes of Binary Solution UF of Cl−/ H2PO4− at pH 5.4 as a Function of Bilayer Number and NaCl Concentration in the Dipping Solution (0−1.0 M)

(5)

permeate flux (m3/m2 day)

The positively charged polymer unit here is PEI which would prefer phosphate. Chloride will be replaced from the matrix increasing its concentration in permeate. As more and more chloride is available in the matrix with more salt in the depositing PE, more Cl− appeared in permeate (Scheme 1). To further prove chloride exclusion, the feed containing NaNO3 instead of NaCl was ultrafiltered. The results are presented in Table 3. Chloride ions again showed permeate enrichment.

no. of bilayers

0M

0.1 M

0.2 M

0.3 M

0.4 M

0.5 M

1.0 M

6 7 8 9

29.56 27.24 25.71 19.57

29.23 26.24 23.35 18.04

28.03 25.91 23.39 19.09

27.21 24.67 22.85 19.17

20.17 18.95 19.11 12.10

19.89 18.71 18.59 13.05

17.67 15.39 13.89 13.17

Table 3. Percent Rejections and Selectivities of Binary Solution UF of NO3−/H2PO4− at pH 5.4 as a Function of Bilayer Number (Concentration of NaCl = 0.5 M) no. of bilayers

flux (m3/m2 day)

nitrate rejection (%)


chloride rejection (%)

6 7 8 9

19.05 17.75 17.09 14.10

1.33 9.73 15.82 15.30

36.25 59.81 61.66 73.05

−80.59 −200.45 −280.30 −346.90

Thus by modulating the polyelectrolyte construction by ionic strength we could bring the chloride to permeate giving an overall high selectivity for the chosen ion pair. Filtration experiments were also carried out in the presence of humic acid. The results are presented in Table 4. The phosphate/ chloride trend was almost similar. The presence of humic acid did not alter this pattern. If one looks at the flux given in Table 5 for a particular bilayer, it decreases as expected. An increase in trend is expected for an increase in build up salt as the possibility for more open structure increases. Though a large flux is obtained a reverse in tendency is observed. Since PEI, the cationic layer, has an affinity toward phosphate, it can strongly bind and reduce permeate flux. This result is agreeable with the observations of Vankelecom et al. for SPEEK/PDDA where phosphate showed an intermolecular bonding with positively charged polymer unit.16 As a matter of fact, several experimental parameters can affect the construction of multilayers and hence the permselectivity. The influence of polyelectrolyte solution concentration on transport properties was studied using multilayers deposited from three different concentrations at 0.01, 0.02, and 0.03 M. The results are presented in Figure 5. At a concentration of 0.01 M polyelectrolyte in 0.5 M NaCl, there is appreciable phosphate rejection which increases from 37.3% for (PEI/ PSS)6 to 82.3% for (PEI/PSS)9. Correspondingly the Cl−/ H2PO4− selectivity increased from 3.7 to 20.3. For a concentration of 0.03 M, a rejection percentage of 88% with a Cl−/H2PO4− selectivity is 29 are observed. However, the flux is rather low. It is clear that 0.02 M polyelectrolyte in 0.5 M

Figure 5. Percent rejection of H2PO4− against bilayer number at different polyelectrolyte concentration.

NaCl is best suited for the selective recovery of phosphate from a chloride containing solution in the studied concentration regime of polyelectrolytes. The character of the surrounding medium to which the bilayer is exposed can also alter the filtration pattern due to changes that can take place to the bl. In addition to PEI being a weak base polymer, the multilayer will respond to the changes in pH. The phosphate form in solution state changes from protonated to deprotonated as we go from acidic to basic pH (H3PO4 to PO43‑).15 In order to study the effect of feed pH on rejection pattern, experiments were conducted at two different pHs, 5.4 and 10, and the results are presented in Table 6. At pH 5.4 for the PEI/PSS multilayer prepared from 0.3 M NaCl, rejection of phosphate through seven bl was 51.13 with a selectivity of 4.22, whereas at pH 10, the observed rejection was 0.88% with a selectivity of 1.13. Tuning the solution pH is often used as a trigger to control the permeability in applications related to drug delivery. pH dependent open and close structure of multilayers are widely discussed for loading and release of dye molecules. Similarly the instability of bilayers at high pH is also known.38 The observed decrease in rejection at alkaline pH may be due to the dislocation of bilayer which is

Table 4. Water Flux, Percent Rejections, and Solution Fluxes of Binary Solution Ultrafiltration of Humic Acid/H2PO4− through PEI/PSS on Nylon (Concentration of NaCl = 0.5 M) no. of bilayers

water flux (m3/m2 day)

humic acid rejection (%)



flux (m3/m2 day)

unmodified nylon 6 7 8 9

73.50 59.96 57.06 55.35 49.79

6.25 30.32 34.32 36.54 37.09

−1.05 86.15 92.96 97.05 98.11

−199.17 −235.41 −252.39 −284.81

65.06 11.05 10.75 7.99 5.10

12749


Langmuir

Article

Table 6. Rejections, Solution Fluxes, and Selectivities of Binary Salt UF Experiments of Chloride and Phosphate through (PEI/PSS)n Coated Nylon Membrane (Concentration of NaCl = 0.3 M) no. of bilayers pH of feed solution = 5.4 7 8 pH of feed solution = 10 7 8




flux (m3/m2 day)

−106.33 −114.03

51.13 55.37

4.22 4.79

25.17 24.85

−12.14 −88.66

0.88 12.53

1.13 2.16

29.15 28.24

Table 7. Rejections, Solution Fluxes, and Selectivities of Binary Salt UF Experiments of Chloride and Phosphate through (PEI/PSS)n Coated Nylon and Polyethersulfone Membranes (Concentration of NaCl = 1 M) no. of bilayers 6 bl on nylon 6 bl on PES 7 bl on nylon 7 bl on PES




flux (m3/m2 day)

−183.21

77.56

12.62

17.670

−378.02

48.64

9.31

0.935

−220.14

84.76

21.01

15.390

−407.17

62.07

13.37

0.876

ment is observed in either case as well as rejection of phosphate though to a lesser extent in PES. Moreover the substrates used had the same pore size unlike the findings of Liu et al.40 More investigations regarding this aspect are in progress. The stability of PEI/PSS films in external salt solutions of high ionic strength was investigated. For this study, the films (7 and 8 bilayers prepared from 0.5 M NaCl solutions) were exposed to 3 and 5 M NaCl concentrations. Recently Han et al. pointed out that post treatment of films with solutions of high ionic strength may result in mass loss.41 After the post treatment, the respective membranes were used for the ultrafiltration of binary electrolytes. The corresponding results are presented in Table 8. The percentage rejection of H2PO4−

evident from the SEM image (Figure 6). The film morphology looks rather distinct, with the membrane exposed to higher pH appearing thinner. The high pH of feed which is different from the solution pH of PEI may cause instability of bilayer that leads to drastic drop in phosphate rejection. For confirming our observation we did same experiment through eight bilayers and here too rejection was found to be as low as 12.5. It is a known fact that, due to the acid−base equilibria in weak polyelectrolytes, multilayer films respond to pH changes in the outer solution.39 Besides giving mechanical strength, the support could also play a role in transport profile. The support material can altogether influence the multilayers, their filtration capacity, and their binding.40 We made an attempt to study the role of support in ion rejection of phosphate through the PEI/PSS system. The results are shown in Table 7. Nylon and polyether sulfone (PES) were selected as supports, and for (PEI/PSS)6 in 1.0 M NaCl on nylon, the rejection rate of phosphate was 77.56% and on PES it was 48.64%. The corresponding values of selectivities were 12.62 and 9.3. Permeate flux values also differ much. Flux value was as high as 17.67 for nylon, whereas for PES, flux value was 0.935 for nylon. For (PEI/PSS)7 too, a similar trend was observed. In the case of (PEI/PSS)7 in 1.0 M NaCl on nylon, the rejection rate of phosphate was 84.76%, and on PES it was 62.07%. The corresponding values of selectivities were 21.01 and 13.37 with a flux of 15.39 and 0.876. So the difference in the chemical structures of the supports affects the permeate fluxes significantly. When PES is used as the support, the pores are perhaps covered which seems to be the most probable reason for the decreased flux. As such there is no difference in the transport pattern because permeate enrich-

Table 8. Percent Rejections and Selectvities of UF of Cl−/ H2PO4− through PEI/PSS Multilayers after Annealing in 3 and 5 M NaCl no. of bilayers




flux (m3/m2 day)

7 bl (0 M NaCl) 7 bl (3 M NaCl) 7 bl (5 M NaCl) 8 bl (0 M NaCl) 8 bl (3 M NaCl) 8 bl (5 M NaCl)

−212.34

81.89

17.25

18.71

−62.85

53.72

3.52

26.98

−89.69

52.59

4.00

30.54

−219.25

96.58

93.35

18.59

−85.06

67.59

5.71

26.10

−97.15

45.45

3.61

31.45

Figure 6. SEM image of eight bilayer (PEI-PSS) after UF at pH 5.4 (A) and 10 (B). 12750


Langmuir

Article

as well as the negative solute rejection of Cl− were reduced and hence the respective selectivity too. At the same time, flux values increased indicating a reduction in thickness. The results indicate the possibility of mass loss due to disassembly of the multilayers. At lower salt concentration (3 M), PEI is likely to be dissolved, and the same holds for PSS at 5 M salt concentration.

(2) Elser, J. J. Phosphorus: a limiting nutrient for humanity? Curr. Opin. Biotechnol. 2012, 23, 1−6. (3) Baker, L. A. Can urban P conservation help to prevent the brown devolution? Chemosphere 2011, 84, 779−784. (4) Cordell, D.; Rosemarin, A.; Schröder, J. J.; Smit, A. L. Towards global phosphorus security: A systems framework for phosphorus recovery and reuse options. Chemosphere 2011, 84, 747−758. (5) Ashley, K.; Cordell, D.; Mavinic, D. A brief history of phosphorus: From the philosopher’s stone to nutrient recovery and reuse. Chemosphere 2011, 84, 737−746. (6) Van Vuuren, D. P.; Bouwmana, A. F.; Beusen, A. H. W. Phosphorus demand for the 1970 − 2100 period: A scenario analysis of resource depletion. Global Environ.Chang. 2010, 20, 428−439. (7) Vaccari, D. A. Phosphorus: A Looming Crisis. Sci. Am. 2009, 300, 54−59. (8) Rittmann, B. E.; Mayer, B.; Westerhoff, P.; Edwards, M. Capturing the lost phosphorus. Chemosphere 2011, 84, 846−853. (9) Mir, F. Q.; Shukla, A. Negative rejection of NaCl in ultrafiltration of aqueous solution of NaCl and KCl using sodalte octahydrate zeolite-clay charged ultrafiltration membrane. Ind. Eng. Chem. Res. 2010, 49, 6539−6546. (10) Decher, G. Fuzzy nanoassemblies: toward layered polymeric multicomposites. Science 1997, 277, 1232−1237. (11) Harris, J. J.; Stair, J. L.; Bruening, M. L. Layered polyelectrolyte films as selective, ultrathin barriers for anion transport. Chem. Mater. 2000, 12, 1941−1946. (12) Dai; Dai, J.; Jenson, A. W.; Mohanty, D. K.; Erndt, J.; Bruening, M. L.; et al. Controlling the permeability of multilayered polyelectrolyte films through derivatisation, crosslinking, and hydrolysis. Langmuir 2001, 17, 931−937. (13) Stanton, B. W.; Harris, J. J.; Miller, M. D.; Bruening, M. L. Ultrathin, multilayered polyelectrolyte films as nanofiltration membranes. Langmuir 2003, 19, 7038−7042. (14) Liu, X.; Bruening, M. L. Size selective transport of uncharged solutes through multilayered polyelectrolyte membranes. Chem. Mater. 2004, 16, 351−357. (15) Hong, S. U.; Ouyang, L.; Bruening, M. L. Recovery of phosphate using multilayer polyelectrolyte nanofiltration membranes. J. Membr. Sci. 2009, 327, 2−5. (16) Ahmmadiannamini, P.; Li, X.; Goyens, W.; Meesschaert, B.; Vankelecom, I. J. Multilayered PEC nanofiltration membranes based on SPEEK/PDDA for anion separation. J. Membr. Sci. 2010, 360, 250−258. (17) Baburaj, M. S.; Aravindakumar, C. T.; Sreedhanya, S; Thomas, A. P.; Aravind, U. K. Treatment of model effluents with PAA/CHI and PAA/PEI composite membranes. Desalination 2012, 288, 72−79. (18) Aravind, U. K.; George, B.; Baburaj, M. S.; Thomas, S.; Thomas, A. P.; Aravindakumar, C. T. Treatment of industrial effluents using polyelectrolyte membranes. Desalination 2010, 252, 27−32. (19) Aravind, U. K.; Mathew, J.; Aravindakumar, C. T. Transport studies of BSA, lysosyme and ovalbumin through chitosan/polystyrene sulfonate multilayer membrane. J. Membr. Sci. 2007, 299, 146−155. (20) Mathew, J.; Aravindakumar, C. T.; Aravind, U. K. Effect of ionic strength and protein concentration on the transport of proteins through chitosan/polystyrene sulfonate multilayer membrane. J. Membr. Sci. 2008, 325, 625−632. (21) Birnbaum, E. R.; Rau, K. C.; Sauer, N. N. Selective anion binding from water using soluble polymers. Sep. Sci. Technol. 2003, 38, 389−404. (22) Elzbieciak, M.; Zapotoczny, S.; Nowak, P.; Krastev, R.; Nowakowska, M.; Warszynski, P. Influence of pH on the structure of multilayer films composed of strong and weak polyelectrolytes. Langmuir 2009, 25, 3255−3259. (23) Krasemann, L.; Tieke, B. Selective ion transport across selfassembled alternating multilayers of cationic and anionic polyelectrolytes. Langmuir 2000, 16, 287−290. (24) Lewis, S. R.; Datta, S.; Gui, M.; Coker, E. L.; Huggins, F. E.; Daunert, S.; Bachas, L.; Bhattacharyya, D. Reactive nanostructured

4. CONCLUSIONS Layer by layer assembly of PEI/PSS films on a porous nylon substrate results in high flux low pressure filtration membranes that selectively recover phosphate (H2PO4−) in the presence of chloride. The membranes showed negative solute rejection for chloride which in turn facilitated selective removal of phosphate. The presence of supporting salt in deposited polyelectrolyte solution controlled the film morphology. The diffusion potential created by the difference in concentration of Na+ and Cl− in feed as well as the competition between chloride and phosphate for positively charged polymer unit resulted in permeate enrichment (negative solute rejection) of chloride. The affinity of PEI for phosphate too resulted in ∼98% rejection of phosphate. This is a very encouraging result as such a high selectivity and rejection of H2PO4− at low pressure end has never been reported before. The transport characteristics were dependent on the concentration of polyelectrolyte solutions, pH of the feed solution, and the nature of the support. Negative solute rejection (which can be modulated by the concentration of supporting salt) of chloride and high rejection of phosphate at pH 5.4 at remarkably low pressure make this membrane highly attractive for the removal of phosphate from the aqueous system at ppm level. Desorption possibilities of phosphate are yet another area which has to be worked out. The most attractive feature of the present work is the use of low pressure end filtration (UF) for the separation of the Cl−/H2PO4− pair. Ultrafiltration/microfiltration is being considered as one of the most favored low pressure driven technologies for water purification/treatment today.

■

ASSOCIATED CONTENT

S Supporting Information *

Additional Tables 1−3 and Figures 1−7. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Tel.: +91-491-2732120. Fax: +91-481-2731009. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS D.V.J. is thankful to University Grants Commission (Delhi) for the financial support under FIP scheme. U.K.A. and C.T.A. are thankful to DST (Women Scientist Project) and DST-Purse, respectively, for financial support.

■

REFERENCES

(1) Cordell, D; Drangert, J.-O.; White, S. The story of phosphorus: Global: food security and food for thought. Global Environ. Chang. 2009, 19, 292−305. 12751


Langmuir

Article

membranes for water purification. Proc. Natl. Acad. Sci. 2011, 108, 8577−8582. (25) Shannon, M. A.; Bohn, P. W.; Elimelech, M.; Georgiadis, J. G.; Marinas, B. J.; Mayes, A. M. Science and technology for water purification in the coming decades. Nature 2008, 452, 301−310. (26) Picart, C.; Lavalle, P.; Hubert, P.; Cuisinier, F. J. G.; Decher, G.; Schaaf, P.; Voegel, J.-C. Buildup mechanism for Poly(L-lysine)/ Hyaluronic acid films onto a solid surface. Langmuir 2001, 17, 7414− 7424. (27) Picart, C; Mutterer, J.; Richert, L.; Luo, Y.; Prestwich, G. D.; Schaaf, P.; Voegel, J.-C.; Lavalle, P Molecular basis for the explanation of the exponential growth of polyelectrolyte multilayers. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 12531−12535. (28) McAloney, R. A.; Sinyor, M.; Dudnik, V.; Goh, M. C. Atomic force microscopy studies of salt effects on polyelectrolyte multilayer film morphology. Langmuir 2001, 17, 6655−6663. (29) Schlenoff, J. B.; Ly, H.; Li, M. Charge and mass balance in polyelectrolyte multilayers. J. Am. Chem. Soc. 1998, 120, 7626−7634. (30) Farhat, T. R.; Schlenoff, J. B. Ion transport and equilibria in polyelectrolyte multilayers. Langmuir 2001, 17, 1184−1192. (31) Farhat, T. R.; Schlenoff, J. B. Doping controlled ion-diffusion in polyelectrolyte multilayers: mass transport in reluctant exchangers. J .Am. Chem. Soc. 2003, 125, 4627−4636. (32) Ball, V; Hubsch, E; Schweiss, R.; Voegel, J.-C.; Schaaf, P.; Knoll, W. Interactions between Multivalent Ions and ExponentiallyGrowing Multilayers: Dissolution and Exchange Processes. Langmuir 2005, 21, 8526−8531. (33) Lavalle, P.; Voegel, J.-C.; Vautier, D.; Senger, B.; Schaaf, P.; Ball, V. Dynamic aspects of films prepared by a sequential deposition of species: perspectives for smart and responsive materials. Adv. Mater. 2011, 23, 1191−1221. (34) El Haitami, A. E.; Martel, D.; Ball, V.; Nguyen, H. C.; Gonthier, E.; Labbe, P.; Voegel, J. C.; Schaff, P.; Senger, B.; Boulmedais, F. Effect of supporting electrolyte anion on the thickness of PSS/PAH multilayer films and on their permeability to an electroactive probe. Langmuir 2009, 25, 2282−2289. (35) Bardot, C.; Gaubert, E.; Yaroschuk, A. E. Unusual mutual influence of electrolytes during pressure −driven transport of their mixtures across charged porous membranes. J. Membr. Sci. 1995, 103, 11−17. (36) Duval, J. F. L.; Kuttner, D.; Werner, C.; Zimmermann, R. Electrohydrodynamics of soft polyelectrolyte multilayers: point of zero-streaming current. Langmuir 2011, 27, 10739−10752. (37) Childress, A. E.; Elimelech, M. Relating nanofiltration membrane performance to membrane charge (electrokinetic) characteristics. Environ. Sci. Technol. 2004, 34, 3710−3716. (38) Burke, S. E.; Barrett, C. J. pH dependent loading and release behaviour of small hydrophilic molecules in weak polyelectrolyte multilayer films. Macromolecules 2004, 37, 5375−5384. (39) Tagliazucchi, M.; Williams, F. J.; Calvo, E. J. Effect of acid-base equilibria on the donnan potential of layer-by-layer redox polyelectrolyte multilayers. J. Phys. Chem. B 2007, 111, 8105−8113. (40) Liu, G.; Dotzauer, D. M.; Bruening, M. L. Ion-exchange membranes prepared using layer by layer polyelectrolyte deposition. J. Membr. Sci. 2010, 354, 198−205. (41) Han, L.; Mao, Z.; Wuliyasu, H.; Wu, J.; Gong, X.; Yang, Y.; Gao, C. Modulating the structure and properties of Poly(sodium 4styrenesulfonate)/Poly(diallyldimethylammonium chloride) multilayers with concentrated salt solutions. Langmuir 2012, 28, 193−199. (42) Raton, B. Handbook of Chemistry and Physics; CRC Press: Boca Raton, FL, 2000.

12752


Phosphate Recovery by High Flux Low Pressure Multilayer Membranes

Recommend Documents